The most consequential change in the UK’s next Research Excellence Framework is that outputs will no longer follow researchers when they move institutions in the way they did under REF 2021. This page covers the confirmed REF 2029 output portability rules — the “decoupling” of outputs from individual staff, the “substantive link” test that replaces the old portability rule, and the narrower exception that still applies to long-form and extended-process outputs such as monographs. It is distinct from the general REF 2029 overview and from the REF 2029 output eligibility period guide, which covers a related but separate question — the date range within which an output must first have been made publicly available, not whether it travels with a researcher who changes jobs.
What ‘portability’ meant under REF 2021, and what changes
Under REF 2021 and earlier exercises, an output was generally submissible by the institution employing the author on the census date, regardless of which institution the underlying research had actually been conducted at — in effect, outputs were “portable”: a researcher who changed jobs could usually bring their published outputs with them for submission by their new employer. This made outputs, in practice, an attribute of the individual researcher rather than the institution.
REF 2029 confirms a structural break from that approach. Section 4 of REF 2029’s published guidance, covering Contributions to Knowledge and Understanding (CKU) — REF 2029’s name for what REF 2021 called the Outputs element — introduces decoupling: institutions no longer submit a list of staff paired with the outputs attributed to them. Instead, an institution submits outputs, and each output must independently satisfy a substantive link requirement connecting it to that institution. REF 2029’s own guidance frames the rationale as proportionality and flexibility — keeping the size of an institution’s submission tied to the scale of its actual research activity, and giving institutions more room to represent a diverse range of research and researchers — rather than as a mechanism to track a fixed list of “REF-returnable” staff.
The substantive link test
A substantive link is REF 2029’s replacement for the old portability rule. It describes the relationship between a submitting institution and a specific output, and it must be evidenced through an eligible employment relationship between the institution and an author (or equivalent contributor) who made a significant research contribution to that output. Two conditions define an eligible employment relationship under the published guidance:
- Duration and intensity: a minimum of 0.2 FTE, sustained for at least 12 months of continuous employment with the submitting institution.
- Job expectations: the role’s descriptor must include an explicit expectation of research activity, consistent with REF’s definition of research.
Critically, that employment relationship does not have to be current at the point of publication — it can exist either at the point the output is first made publicly available, or during a defined lookback window beforehand. For standard outputs, the lookback window is two years prior to publication. For long-form and extended-process outputs, discussed below, the window extends to five years.
The long-form and extended-process exception
Decoupling removes general output portability, but REF 2029’s guidance carves out a narrower, evidenced exception for output types that take substantially longer to produce than a typical journal article — recognising that a rigid two-year lookback window would otherwise penalise the research most likely to span an institutional move. The published guidance defines two categories that qualify:
- Long-form outputs — outputs of extended scale or scope, the paradigm case being a monograph.
- Extended-process outputs — outputs representing the culmination of a substantial, discrete research endeavour involving intense activity sustained over a period of time, usually five years or longer.
For these two categories, a researcher employed at an institution at the point that institution makes its REF submission may still submit an output published before their employment there began, provided both of the following hold:
- The output was first made publicly available within the REF 2029 output eligibility period (1 January 2021 – 31 December 2028); and
- The employment relationship with the submitting institution commenced within five years of the output first being made publicly available.
Using this exception is not automatic: the published guidance requires a 100-word justification submitted alongside the output, subject to panel assessment, rather than a simple date check.
What stays fully portable
Decoupling applies to REF submission eligibility specifically — it does not touch a researcher’s ownership of, or ability to use, their own work. REF 2029’s guidance is explicit that researchers retain their CV, their publications, and their expertise, and that these remain fully portable in every professional and academic sense for job applications, promotion, and grant funding applications. What changes is narrower and more specific: which institution may list a given output in its own REF submission, and therefore which institution’s REF outcome that output contributes to.
Why this matters for staff mobility
Because REF outcomes drive a portion of UK quality-related (QR) research funding, and because REF performance factors into hiring, promotion, and institutional reputation, the removal of general output portability has been one of the most contested elements of the REF 2029 reform process. Sector commentary — including reporting in Times Higher Education on the decision to proceed with decoupling despite objections raised during consultation — has raised concern that researchers who move institutions could find outputs produced under a previous employer effectively “retained” by that employer for REF purposes, outside the narrow long-form/extended-process exception, potentially affecting how institutions value hiring a researcher whose recent outputs would not travel with them. REF 2029’s own published rationale is that funding tied to REF performance should follow the institution that substantively enabled the research environment in which an output was produced, rather than following an individual regardless of where that enabling work actually happened. Research administrators supporting recruitment, retention, or early-career researchers changing institutions should treat the substantive-link and lookback-window mechanics above as directly relevant to REF planning conversations, not just a compliance footnote.
Confirmed vs. still settling
Decoupling and the substantive-link framework are published, confirmed elements of REF 2029’s Section 4 (CKU) guidance, not a proposal still under open consultation — the core policy direction was confirmed to proceed following the sector consultation period. That said, REF 2029’s guidance site states that all REF 2029 guidance modules will be formally finalised in 2026 and “may be subject to small revisions” in the interim, and the site maintains a change log tracking exactly this kind of update. Detailed audit and submission-process guidance for how substantive-link evidence will actually be checked has not been fully published as of this writing. Research administrators should treat the mechanics described on this page — the FTE/duration thresholds, the two- and five-year lookback windows, and the long-form/extended-process exception — as the current published position, and check the guidance change log at 2029.ref.ac.uk directly before relying on any one summary, including this one, as the final word.
Frequently asked questions
Can a researcher still be submitted for REF 2029 by their new institution after moving jobs?
The researcher isn’t submitted at all under REF 2029’s decoupled model — institutions submit outputs, not staff-output pairs. Whether a specific output an author contributed to can be submitted by their new institution depends on whether that institution can evidence a substantive link to the output (an eligible employment relationship at publication or within the relevant lookback window), not on where the author currently works.
Does the portability change affect a researcher’s CV or job applications?
No. REF 2029’s guidance is explicit that a researcher’s publications, CV, and expertise remain fully portable for hiring, promotion, and grant applications. Decoupling only affects which institution may include a given output in its own REF submission.
What counts as a ‘long-form’ or ‘extended-process’ output for the five-year exception?
Long-form outputs are defined as outputs of extended scale or scope — monographs are the standard example. Extended-process outputs are defined as the culmination of a substantial, discrete research endeavour involving intense, sustained activity, usually of five years or longer. Both require a submitted justification and are subject to panel assessment rather than being accepted automatically.
Is the REF 2029 portability policy final?
The decision to proceed with decoupling and the substantive-link framework is confirmed. However, REF 2029’s guidance programme as a whole is due to be formally finalised in 2026, and REF 2029’s own guidance site notes that modules may still see small revisions before then — check the published change log for updates.
